CN103605114A - Non-broadside array airborne radar short range clutter suppression method based on multiple frequencies - Google Patents

Abstract

The invention discloses a non-broadside array airborne radar short range clutter suppression method based on multiple frequencies. The non-broadside array airborne radar short range clutter suppression method is implement in the following steps: (1) a distance unit range and a Doppler value range of strong short range clutter are acquired; (2) echo data of all distance units received by a radar are subjected to pitching sum and difference beam forming to obtain sum and difference beam data; (3) the sum and difference beam data within the distance unit range of the strong short range clutter are subjected to smoothing to obtain a smoothing result of the distance units within the distance unit range; (4) the sum and difference beam data beyond the distance unit range are subjected to pitching beam forming to obtain a smoothing result of the distance units beyond the distance unit range; (5) the smoothing result in the step (3) and the smoothing result in the step (4) are integrated to obtain a smoothing result of all the distance units, and then short range clutter suppression is completed. By means of the non-broadside array airborne radar short range clutter suppression method based on the multiple frequencies, the distance stability of a radar echo is improved, the clutter suppression performance of a radar system is improved, the amount of calculation is lowered, and the non-broadside array airborne radar short range clutter suppression method can be applied to target detection.

Description

Non-positive side battle array airborne radar short range clutter suppression method based on multiple frequency

Technical field

The invention belongs to signal processing technology field, further relate to radar clutter suppression method, the short range clutter distributing in the space-time two-dimensional plane that can be used for suppressing to cause when airborne radar antenna non-working side is settled.

Background technology

In airborne radar signal is processed, because the irradiation of radar signal main lobe can produce very strong land clutter, especially, when airborne radar antenna non-working side is settled, the Range-dependent that clutter presents in space-time two-dimensional distribution can cause stronger short range clutter.Specifically, between each range unit of short range, luffing angle changes greatly, cause the Doppler frequency of clutter can produce along with the variation of range unit comparatively strong variation, so just produced short range clutter, the destruction that the appearance of short range clutter is serious the stationarity of clutter, make to obtain abundant independent same distribution sample, cause traditional space-time adaptive to process the clutter rejection degradation of (STAP) technology.

The method of eliminating at present clutter Range-dependent mainly contains the full spectral domain algorithm of compensation class algorithm, change of scale class algorithm, interpolation class.

Compensation class algorithm is that each training sample is compensated respectively, and it mainly comprises: the Doppler effect correction that calendar year 2001 Kreyenkamp O and Klemm R propose; Calendar year 2001 Pearson F and Borsari G propose high-order Doppler effect correction; The self-adaptation angle Doppler effect correction that Melvin W and Davis M in 2007 proposes on periodical Aerospace and Electronic Systems; The angle Doppler effect correction that Colone F in 2011 proposes, this class algorithm all can compensate clutter Range-dependent at clutter without range ambiguity in the situation that effectively, but when clutter exists range ambiguity, this class algorithm is no longer applicable;

2011 by Ries P, the method that Lapierre F D and Verly J G adopts spectrum to estimate at the upper change of scale class algorithm proposing of periodical Aerospace and Electronic Systems (AES), the clutter covariance matrix of reconstruct unit to be detected, then according to reconstruct clutter covariance matrix out, the clutter in training sample is transformed in the clutter space of sample to be detected, this class algorithm can effectively compensate the Range-dependent of clutter when clutter exists range ambiguity, but this class algorithm operation quantity is larger, is not easy to practical application;

The full spectral domain algorithm of the interpolation class that Liu Jinhui etc. propose in electronics and information journal adopts least square technology the clutter of training sample to be transformed in the clutter space of sample to be detected, also can effectively compensate the Range-dependent of clutter, but steering vector in the time of need to estimating clutter empty, but due to the reason of error estimate empty time steering vector and actual when empty steering vector have the deviation that can not estimate, so this class algorithm clutter rejection when there is array error can decline greatly.

Summary of the invention

The object of the invention is to the deficiency for above-mentioned prior art, a kind of non-positive side battle array airborne radar short range clutter suppression method based on multiple frequency has been proposed, to avoid the Range-dependent of training sample, improve the distance stationarity of radar return, improve the clutter rejection of radar system, reduce operand.

Technical scheme of the present invention is achieved in that

One. know-why

For airborne radar, normally by launching continuously the waveform of several groups of different pulse repetition, solve the problem that target range is fuzzy, in certain time interval, because the short range clutter of same distance unit in these several groups of different pulse repetition echoes has higher correlativity, therefore can estimate the covariance matrix of clutter as training sample by the data of same distance unit in these several groups of different pulse repetition echoes.For the less problem of training sample, can, by adopting pitching and the poor adaptive processing technique of two-freedom, reduce the needed training sample number of accurate estimation clutter covariance matrix, by being filtered the inhibition of non-positive side battle array airborne radar clutter in pairs.

Two. technical scheme

The technical step of realizing the object of the invention comprises as follows:

(1) obtain range unit scope and Doppler's span of the strong clutter of short range:

1a) according to pulse repetition rate maximum in several pulse repetition raties of radar emission and the flying height of bandwidth and radar, calculate the not numerical value of fuzzy distance unit, ultimate range unit numerical value L as the strong clutter of short range, and calculating altitude line range unit H, the range unit scope that obtains the strong clutter of short range is H～L;

1b) calculate in the range unit scope H～L of the strong clutter of short range Doppler's span of the strong clutter of short range in each range unit main cone face;

(2) echo data of all range units of I group different pulse repetition radar being received carries out pitching and difference beam forms, obtain all pulse repetition raties all range units to be processed with difference beam data X ₁, X ₂... X _i... X _i, i=1 wherein, 2...I, I is the number of the data of the different pulse repetition that receives of radar;

(3) to carrying out filtering with difference beam data in the range unit scope H～L of the strong clutter of short range:

3a) from all range units to be processed with difference beam data X ₁, X ₂... X _i... X _iin, the data x of l range unit within the scope of taking-up H～L range unit _{1, l}, x _{2, l}... x _i,l... x _i,l;

3b) from step 1b) take out the doppler values in the middle of being positioned in Doppler's span of the strong clutter of short range of l range unit calculating, with this doppler values formation time domain steering vector v _{1, l}, v _{2, l}... v _i,l... v _i,l, then use this time domain steering vector respectively with the data x of l range unit _{1, l}, x _{2, l}... x _i,l... x _i,lmultiply each other, by the result multiplying each other, carry out the covariance matrix of short range clutter, then with the covariance matrix of this short range clutter, carry out the calculating of adaptive weight;

3c) use the data x of weights to l range unit _{1, l}, x _{2, l}... x _i,l... x _i,lcarry out filtering, obtain the filtering result of the data of l range unit: y _{1, l}, y _{2, l}... y _i,l... y _i,l;

3d) repeat 3b)-3c), until obtain all range units with difference beam data X ₁, X ₂... X _i... X _ibe positioned at the data filtering result of H～L range unit scope;

(4) to outside the range unit scope H～L of the strong clutter of short range to carry out pitching wave beam synthetic with difference beam data:

4a) from i all range units of pulse repetition rate to be processed with difference beam data X _ithe data of the extraneous range unit of middle taking-up H～L range unit, carry out pitching wave beam synthetic, obtain i pulse repetition rate with difference beam data X _iin be positioned at the filtering result of the extraneous range unit of H～L range unit, i=1 wherein, 2...I, I is the number of the data of the different pulse repetition that receives of radar;

4b) repeat 4a), until obtain all pulse repetition raties with difference beam data X ₁, X ₂... X _i... X _ibe positioned at the filtering result of the extraneous data of H～L range unit;

(5) integral data, completes short range clutter and suppresses:

5a) get the filtering result of i all range units of pulse repetition rate in step 3) and step 4), according to range unit sequence number, carry out Data Integration, obtain the integration kalman filter result Y of i pulse repetition rate _i;

5b) repeating step 5a), obtain the integration kalman filter result Y of all pulse repetition raties ₁, Y ₂... Y _i... Y _i, complete the inhibition to airborne radar short range clutter.

The present invention compared with prior art, has the following advantages:

1. the present invention compares along the method for choosing sample apart from dimension with common, because the echo data of choosing a plurality of pulse repetition rate same distance unit is as sample, can avoid the Range-dependent problem of training sample, can estimate comparatively exactly clutter covariance matrix, improve the distance stationarity of radar return, improve the clutter rejection of radar system.

2. the present invention is owing to adopting pitching and the poor adaptive processing technique of two-freedom to suppress the strong clutter of short range, can reduce the requirement to training sample number, reduce operand simultaneously, the difficulty of estimate covariance is further reduced, better suppressed short range clutter.

Accompanying drawing explanation

Fig. 1 is realization flow figure of the present invention;

Fig. 2 is not for being used the range Doppler image of partial distance unit of the radar return of the inventive method;

Fig. 3 is for being used the inventive method to process the range Doppler image of the partial distance unit of radar return later;

Fig. 4 is the partial distance cell power figure of target place Doppler's passage from Fig. 2 and Fig. 3.

Embodiment

With reference to Fig. 1, implementation step of the present invention is as follows:

Step 1, obtains range unit scope and Doppler's span of the strong clutter of short range.

1a) according to pulse repetition rate f maximum in several pulse repetition raties of radar emission _maxwith the flying height h of bandwidth B and radar, calculate the not numerical value of fuzzy distance unit

as the ultimate range unit numerical value L of the strong clutter of short range, and according to altitude line range unit H, the range unit scope that obtains the strong clutter of short range is H～L;

The strong clutter of short range that 1b) calculates l range unit in the strong clutter range unit of short range scope H～L is at pitching main cone face interior orientation angle θ _lspan be:

${\mathrm{\θ}}_l\∈[\arccos \left(\frac{\cos ({\mathrm{\θ}}_0-0.5\×{\mathrm{\θ}}_{3\mathrm{dB}})\cos {\mathrm{\φ}}_0}{\cos {\mathrm{\φ}}_l}\right),\arccos \left(\frac{\cos ({\mathrm{\θ}}_0+0.5\×{\mathrm{\θ}}_{3\mathrm{dB}})\cos {\mathrm{\φ}}_0}{\cos {\mathrm{\φ}}_l}\right)],$

Wherein, θ ₀the position angle of radar main beam pointing, φ ₀for the angle of pitch of radar main beam pointing, θ _3dBfor main beam 3dB width,

be l the angle of pitch that range unit is corresponding;

1c) by the strong clutter of short range of l range unit at pitching main cone face interior orientation angle θ _lspan, the span of calculating l the strong clutter Doppler of short range in range unit main cone face is:

$f_{\mathrm{dl}}\∈[\min \left(\frac{2v}{\mathrm{\λ}}\cos ({\mathrm{\θ}}_l-{\mathrm{\θ}}_{\mathrm{\α}})\cos {\mathrm{\φ}}_l\right),\max \left(\frac{2v}{\mathrm{\λ}}\cos ({\mathrm{\θ}}_l-{\mathrm{\θ}}_{\mathrm{\α}})\cos {\mathrm{\φ}}_l\right)],$

Wherein, v is the flying speed of radar, and λ is radar wavelength, φ _lthe angle of pitch of l range unit, θ _αfor the angle of antenna array axially and between carrier aircraft heading;

1d) repeat 1b)-1c), obtain the span of the strong clutter Doppler of short range in all range unit main cone faces in the range unit scope H～L of the strong clutter of short range.

Step 2, the echo data that radar is received carries out pitching and difference beam forms.

The echo data that takes out all range units of I group different pulse repetition from the data that radar receives, carries out pitching and difference beam and forms, obtain all pulse repetition raties all range units to be processed with difference beam data X ₁, X ₂... X _i... X _i, i=1 wherein, 2...I, I is the number of the data of the different pulse repetition that receives of radar;

Step 3, to carrying out filtering with difference beam data in the range unit scope H～L of the strong clutter of short range.

3a) from all range units to be processed with difference beam data X ₁, X ₂... X _i... X _iin, the data x of l range unit within the scope of taking-up H～L range unit _{1, l}, x _{2, l}... x _i,l... x _i,l, i=1,2...I;

3b) from step 1b) take out the doppler values in the middle of being positioned in Doppler's span of the strong clutter of short range of l range unit calculating, with this doppler values formation time domain steering vector v _{1, l}, v _{2, l}... v _i,l... v _i,l, each time domain steering vector is: $v_{i,l}={[{1,e}^{j2\mathrm{\π}\×1\×f_l/f_i},e^{j2\mathrm{\π}\×2\×f_l/f_i}...e^{j2\mathrm{\π}\×(K_i-1)\×f_l/f_i}]}^T,i=\mathrm{1,2}...I,$

Wherein, f _ithe size of i pulse repetition rate, f _lthe doppler values that is positioned at centre taking out in Doppler's span of the strong clutter of short range of l range unit determining in step 1, () ^trepresent transposition, K _ithe echo data of value and i pulse repetition rate in coherent pulse number identical, I is the number of the data of the different pulse repetition that receives of radar;

3c) from the data x of l range unit _{1, l}, x _{2, l}... x _i,l... x _i,lthe data x of middle taking-up m array unit _{1, l, m}, x _{2, l, m}... x _{i, l, m}... x _{i, l, m}, m=1 wherein, 2...M, M is the columns of radar array element;

3d) use the data x of the m array unit taking out _{1, l, m}, x _{2, l, m}... x _{i, l, m}... x _{i, l, m}respectively with time domain steering vector v _{1, l}, v _{2, l}... v _i,l... v _i,leach carry out correspondence and multiply each other, obtain the multiplied result c of m array unit _{1, l, m}, c _{2, l, m}... c _{i, l, m}... c _{i, l, m}, c wherein _{i, l, m}=x _{i, l, m}* v _i,l, i=1,2 ... I;

3e) utilize the multiplied result c of m array unit _{1, l, m}, c _{2, l, m}... c _{i, l, m}... c _{i, l, m}, calculate the covariance matrix R of the short range clutter of m array unit _m,l:

$R_{m,l}=\frac{1}{I}\underset{i=1}{\overset{I}{\mathrm{\Σ}}}{c}_{i,l,m}{c}_{i,l,m}^H,$ Wherein, () ^hrepresent conjugate transpose;

3f) take out the covariance matrix R of the short range clutter of m array unit _m,l, utilize following formula to calculate l range unit data x _{1, l}, x _{2, l}... x _i,l... x _i,lin the weight w of m array unit _m,l:

$\left\{\begin{array}{c}\underset{w_{m,l}}{\min }{w}_{m,l}^H{R}_{m,l}{w}_{m,l}\\ s.t.w_{m,l}^H{s}_e=1\end{array}\right.,$ S wherein _e=[1,0] ^t;

3g) take out the weight w of m array unit _m,l, respectively with the data x of m array unit _{1, l, m}, x _{2, l, m}... x _{i, l, m}... x _{i, l, m}multiply each other, obtain m array metadata x _{1, l, m}, x _{2, l, m}... x _{i, l, m}... x _{i, l, m}filtering result y _{1, l, m}, y _{2, l, m}... y _{i, l, m}... y _{i, l, m};

3h) repetition 3c)-3g) obtain the data x of l range unit _{1, l}, x _{2, l}... x _i,l... x _i,lthe filtering result of all M row: y _{1, l}, y _{2, l}... y _i,l... y _i,l;

3i) repeat 3a)-3h), until obtain all range units with difference beam data X ₁, X ₂... X _i... X _ibe positioned at the data filtering result of H～L range unit scope.

Step 4, carries out pitching wave beam to the extraneous and difference beam data of H～L range unit synthetic.

4a) from i all range units of pulse repetition rate to be processed with difference beam data X _ithe data of the extraneous range unit of middle taking-up H～L range unit, carry out pitching wave beam synthetic, obtain i pulse repetition rate with difference beam data X _iin be positioned at the filtering result of the extraneous range unit of H～L range unit, i=1 wherein, 2...I, I is the number of the data of the different pulse repetition that receives of radar;

4b) repeating step 4a), until obtain all pulse repetition raties with difference beam data X ₁, X ₂... X _i... X _ibe positioned at the extraneous data filtering result of H～L range unit.

Step 5, integral data, completes short range clutter and suppresses.

5a) get the filtering result of i all range units of pulse repetition rate in step 3) and step 4), according to range unit sequence number, carry out Data Integration, obtain the integration kalman filter result Y of i pulse repetition rate _i;

5b) repeating step 5a), obtain the integration kalman filter result Y of all pulse repetition raties ₁, Y ₂... Y _i... Y _i, complete the inhibition to airborne radar short range clutter.

Effect of the present invention can be processed and be further illustrated by following emulated data:

1. simulation parameter

Carrier aircraft flying speed 100m/s, flying height 4150m, 32 of coherent pulse numbers, transmitted bandwidth 2.5MHz, operation wavelength 0.1m, 15 of orientation element number of array, 8 of pitching element number of array, array element interval, orientation and pitching array element interval are 0.05m, main beam azimuth 60 degree, the main beam angle of pitch 0 degree, five kinds of pulse repetition raties are used respectively 7337Hz, 6846Hz, 6533Hz, 5979Hz, 5200Hz.

According to systematic parameter emulation the land clutter data of five kinds of pulse repetition raties of forward sight battle array radar, the miscellaneous noise ratio of every array element, every pulse is 50dB.In main beam, add a target, apart from radar 393.48Km, the flying speed of radar is 18m/s relatively, and target is positioned at the 79th range unit, and its Doppler frequency is 360Hz.

2. emulation content and interpretation of result

2.1) echo data of above-mentioned 5 kinds of pulse repetition raties is directly carried out to wave beam and synthesize, then the data that are 7337Hz from pulse repetition rate, take out front 150 range unit data, draw its range Doppler figure, as Fig. 2.

2.2) use the inventive method to process the echo data of 5 kinds of pulse repetition raties, from process data later, take out front 150 range unit data that pulse repetition rate is the data of 7337Hz, draw its range Doppler figure, as Fig. 3.

2.3) from Fig. 2 and Fig. 3, take out respectively the range unit power of 150 points that the Doppler frequency at target place is 360Hz, draw out 150 whole power diagrams that range unit is corresponding, as Fig. 4.

Comparison diagram 2 and Fig. 3, can significantly see and in Fig. 3, use the short range clutter of the inventive method to be inhibited, and is subject to the impact of the strong clutter of short range in Fig. 2, cannot detect target; After short range clutter is inhibited in Fig. 3, can detect target.

As can see from Figure 4, after the inventive method filtering, short range clutter is effectively suppressed, and target reveals completely from clutter.

Claims (5)

1. the non-positive side battle array airborne radar short range clutter suppression method based on multiple frequency, comprises the steps:

(1) obtain range unit scope and Doppler's span of the strong clutter of short range:

1a) according to pulse repetition rate maximum in several pulse repetition raties of radar emission and the flying height of bandwidth and radar, calculate the not numerical value of fuzzy distance unit, ultimate range unit numerical value L as the strong clutter of short range, and calculating altitude line range unit H, the range unit scope that obtains the strong clutter of short range is H～L;

1b) calculate in the range unit scope H～L of the strong clutter of short range Doppler's span of the strong clutter of short range in each range unit main cone face;

(2) echo data of all range units of I group different pulse repetition radar being received carries out pitching and difference beam forms, obtain all pulse repetition raties all range units to be processed with difference beam data X ₁, X ₂... X _i... X _i, i=1 wherein, 2...I, I is the number of the data of the different pulse repetition that receives of radar;

(3) to carrying out filtering with difference beam data in the range unit scope H～L of the strong clutter of short range:

3a) from all range units to be processed with difference beam data X ₁, X ₂... X _i... X _iin, the data x of l range unit within the scope of taking-up H～L range unit _{1, l}, x _{2, l}... x _i,l... x _i,l;

3b) from step 1b) take out the doppler values in the middle of being positioned in Doppler's span of the strong clutter of short range of l range unit calculating, with this doppler values formation time domain steering vector v _{1, l}, v _{2, l}... v _i,l... v _i,l, then use this time domain steering vector respectively with the data x of l range unit _{1, l}, x _{2, l}... x _i,l... x _i,lmultiply each other, by the result multiplying each other, carry out the covariance matrix of short range clutter, then with the covariance matrix of this short range clutter, carry out the calculating of adaptive weight;

3c) use the data x of weights to l range unit _{1, l}, x _{2, l}... x _i,l... x _i,lcarry out filtering, obtain the filtering result of the data of l range unit: y _{1, l}, y _{2, l}... y _i,l... y _i,l;

3d) repeat 3b)-3c), until obtain all range units with difference beam data X ₁, X ₂... X _i... X _ibe positioned at the data filtering result of H～L range unit scope;

(4) to outside the range unit scope H～L of the strong clutter of short range to carry out pitching wave beam synthetic with difference beam data:

4a) from i all range units of pulse repetition rate to be processed with difference beam data X _ithe data of the extraneous range unit of middle taking-up H～L range unit, carry out pitching wave beam synthetic, obtain i pulse repetition rate with difference beam data X _iin be positioned at the filtering result of the extraneous range unit of H～L range unit, i=1 wherein, 2...I, I is the number of the data of the different pulse repetition that receives of radar;

4b) repeat 4a), until obtain all pulse repetition raties with difference beam data X ₁, X ₂... X _i... X _ibe positioned at the filtering result of the extraneous data of H～L range unit;

(5) integral data, completes short range clutter and suppresses:

5a) get the filtering result of i all range units of pulse repetition rate in step 3) and step 4), according to range unit sequence number, carry out Data Integration, obtain the integration kalman filter result Y of i pulse repetition rate _i;

5b) repeating step 5a), obtain the integration kalman filter result Y of all pulse repetition raties ₁, Y ₂... Y _i... Y _i, complete the inhibition to airborne radar short range clutter.

2. the non-positive side battle array airborne radar short range clutter suppression method based on multiple frequency according to claim 1, step 1b wherein) described calculating is in the range unit scope H～L of the strong clutter of short range, Doppler's span of the strong clutter of short range in each range unit main cone face, carry out as follows:

The strong clutter of short range that 1b1) calculates l range unit in the strong clutter range unit of short range scope H～L is at pitching main cone face interior orientation angle θ _lspan be:

${\mathrm{\θ}}_l\∈[\arccos \left(\frac{\cos ({\mathrm{\θ}}_0-0.5\×{\mathrm{\θ}}_{3\mathrm{dB}})\cos {\mathrm{\φ}}_0}{\cos {\mathrm{\φ}}_l}\right),\arccos \left(\frac{\cos ({\mathrm{\θ}}_0+0.5\×{\mathrm{\θ}}_{3\mathrm{dB}})\cos {\mathrm{\φ}}_0}{\cos {\mathrm{\φ}}_l}\right)],$

Wherein, θ ₀the position angle of radar main beam pointing, φ ₀for the angle of pitch of radar main beam pointing, θ _3dBfor main beam 3dB width,

be l the angle of pitch that range unit is corresponding;

1b2) by the strong clutter of short range of l range unit at pitching main cone face interior orientation angle θ _lspan, the span of calculating l the strong clutter Doppler of short range in range unit main cone face is:

$f_{\mathrm{dl}}\∈[\min \left(\frac{2v}{\mathrm{\λ}}\cos ({\mathrm{\θ}}_l-{\mathrm{\θ}}_{\mathrm{\α}})\cos {\mathrm{\φ}}_l\right),\max \left(\frac{2v}{\mathrm{\λ}}\cos ({\mathrm{\θ}}_l-{\mathrm{\θ}}_{\mathrm{\α}})\cos {\mathrm{\φ}}_l\right)],$

Wherein, v is the flying speed of radar, and λ is radar wavelength, φ _lthe angle of pitch of l range unit, θ _αfor the angle of antenna array axially and between carrier aircraft heading.

3. the non-positive side battle array airborne radar short range clutter suppression method based on multiple frequency according to claim 1, wherein step 3b) described in the middle doppler values of use form time domain steering vector v _{1, l}, v _{2, l}... v _i,l... v _i,l, by following formula, undertaken:

$v_{i,l}={[{1,e}^{j2\mathrm{\π}\×1\×f_l/f_i},e^{j2\mathrm{\π}\×2\×f_l/f_i}...e^{j2\mathrm{\π}\×(K_i-1)\×f_l/f_i}]}^T,i=\mathrm{1,2}...I$

Wherein, f _ithe size of i pulse repetition rate, f _lthe doppler values that is positioned at centre taking out in Doppler's span of the strong clutter of short range of l range unit determining in step 1, () ^trepresent transposition, K _ithe echo data of value and i pulse repetition rate in coherent pulse number identical, I is the number of the data of the different pulse repetition that receives of radar.

4. the non-positive side battle array airborne radar short range clutter suppression method based on multiple frequency according to claim 1, wherein said step 3b) with time domain steering vector v _{1, l}, v _{2, l}... v _i,l... v _i,lrespectively with the data x of l range unit _{1, l}, x _{2, l}... x _i,l... x _i,lmultiply each other, by the result multiplying each other, carry out the covariance matrix of short range clutter, carry out as follows:

(3b1) from the data x of l range unit within the scope of H～L range unit _{1, l}, x _{2, l}... x _i,l... x _i,lthe data x of the m array unit of middle taking-up _{1, l, m}, x _{2, l, m}... x _{i, l, m}... x _{i, l, m}, m=1 wherein, 2...M, M is the columns of radar array element;

(3b2) use the data x of the m array unit taking out _{1, l, m}, x _{2, l, m}... x _{i, l, m}... x _{i, l, m}respectively with time domain steering vector v _{1, l}, v _{2, l}... v _i,l... v _i,leach carry out correspondence and multiply each other, obtain the multiplied result c of m array unit _{1, l, m}, c _{2, l, m}... c _{i, l, m}... c _{i, l, m}, c wherein _{i, l, m}=x _{i, l, m}* v _i,l, i=1,2 ... I;

(3b3) utilize the multiplied result c of m array unit _{1, l, m}, c _{2, l, m}... c _{i, l, m}... c _{i, l, m}, calculate the covariance matrix R of the short range clutter of m array unit _m,l:

$R_{m,l}=\frac{1}{I}\underset{i=1}{\overset{I}{\mathrm{\Σ}}}{c}_{i,l,m}{c}_{i,l,m}^H,$

Wherein, () ^hrepresent conjugate transpose, i=1 wherein, 2 ... I, I is the number of the data of the different pulse repetition that receives of radar;

(3b4) repetition (3b1)-(3b3), obtain the data x of l range unit within the scope of H～L range unit _{1, l}, x _{2, l}... x _i,l... x _i,lin the covariance matrix R of short range clutter of all M array unit _{1, l}, R _{2, l}... R _m,l... R _m,l.

5. the non-positive side battle array airborne radar short range clutter suppression method based on multiple frequency according to claim 1, wherein said step 3b), with the covariance matrix of short range clutter, carry out the calculating of adaptive weight, carry out as follows:

(3ba) from the data x of l range unit within the scope of H～L range unit _{1, l}, x _{2, l}... x _i,l... x _i,lthe data x of the m array unit of middle taking-up _{1, l, m}, x _{2, l, m}... x _{i, l, m}... x _{i, l, m}, m=1 wherein, 2...M, M is the columns of radar array element;

(3bb) take out the covariance matrix R of the short range clutter of m array unit _m,l, utilize following formula to calculate l range unit data x _{1, l}, x _{2, l}... x _i,l... x _i,lin the weight w of m array unit _m,l:

$\left\{\begin{array}{c}\underset{w_{m,l}}{\min }{w}_{m,l}^H{R}_{m,l}{w}_{m,l}\\ s.t.w_{m,l}^H{s}_e=1\end{array}\right.,$ S wherein _e=[1,0] ^t;

(3bc) repetition (3ba)-(3bb), obtain the data x of l range unit _{1, l}, x _{2, l}... x _i,l... x _i,lin the weight w of all M array unit _{1, l}, w _{2, l}... w _m,l... w _m,l.

Images